redlewel
I do find it odd that the 2U model has less storage capacity than the 1U on their chart. That doesn't make much sense to me.
The H1B exists to find laborers in areas where we have a shortage of qualified people in US. Given the fact that in many fields we have lots of American graduates that struggle to get jobs, H1B currently has no legitimate purpose in America. It should be abolished entirely.
Its like I'm reading a claude chat response but worse in every way. Its such quantity over quality nonsense. On the off chance it isn't obvious from the get go and you learn its AI written halfway through you feel taken advantage of as you assumed you were reading someone's actual thoughts. But instead you spent several minutes reading something thats just some generic LLM output.
I also can't cleanly nest and stack GUI applications in tmux like I can with TUI programs. I think GUIs are for the most part tailored towards non-power users unless its something like video editing or 3d modeling that is aimed more towards professionals
Much better options out there. Python already has libraries like Pydantic that offload performance to functions written in Rust under the hood.
On HN eventually there isn't much interesting stuff on the front page, on X it pulls you in with great effectiveness
But clearly for work/business requirements using AI is a no brainer for efficiency and speed(which are key concerns for a business). This is under the assumption that a professional is able to review and guide the agentic tooling for a proper output. In order to truly be a professional in terms of software engineering, you need to write a significant amount of programs manually for some time. I don't think that people with 0 prior experience with programming will be effective if handed AI. I use AI for programs for work, and have written manual programs in the past. I want to go back to writing things manually(as to not atrophy important brain regions) so will probably do some personal projects in this manner.
But I don't see a point in writing a book with AI, writing an email with AI, composing a message with AI. Because using AI is leveraging a predictive machine to expand on intent, and the intent in communicating is best done yourself. In my opinion sending AI content as posts, messages, or correspondence is poor application of AI at best, downright disrespectful at worst.
When you have racks that are full of 4U GPU systems that pull 3-6kW each, the power requirements per rack shoot up to 60kW+(with whole rack setups blowing past 100kW). So an AI datacenter requires over 20x the power compared to what they needed in the past. This either strains the power grid near the DC or they need to create on-site power to fulfill demand, which has noise and other concerns.
I work near a datacenter and you can't hear significant noise from the outside, maybe a slight humming sound but quieter than traffic. This is a DC that has been around for over a decade and uses power from the grid since most of the loads are non-GPU based.
If you want LTS(or you can just pin a rolling release around the same build) you can pay for a support contract. Similar to proxmox. The no-subscription product works just fine.
I do use Neovim as an IDE but thats a bit different, as lazynvim does 90% of what I need and the rest is just tiny changes that improves the experience for me. I also use AI to change my lua files simply because its faster and I'd rather focus on programming.
I do use Mikrotik for 2.5g switching though, zero complaints.
Also neovim + claude code + open terminal pane in tmux is the goated combo anyway. Especially more lately you need to understand the code you are writing if you want to do anything important in software, and the best way to do that is neovim :)
Keep learning yung ones